This is a resubmission of a request for funding for the next four years to continue work on carbon dioxide (CO2) challenge studies of patients with panic disorder. In the last grant period we were forced to switch from our originally proposed steady-state canopy method of delivering 5% CO2 to the double breath 35% CO2 method. This occurred because of unexpected changes in the laboratory at Presbyterian Hospital we originally used. Fortunately, the 35% CO2 method yielded very interesting data and, in the meantime, we have installed and perfected a state-of-the art canopy- spirometer-computer system for respiratory studies in our own Biological Studies Unit laboratory. During the last three years many studies from our laboratory and others reinforce the central importance the respiratory disturbance in the genesis of anxiety attacks. The current proposal has five main goals: 1) to determine the dose of CO2 that best distinguishes panic patients from normal controls and social phobics on the basis of anxiogenic response, 2) to study the biological correlates of CO2 induced anxiety by formally measuring physiologic CO2 sensitivity in patients and controls using both the steady-state and the Read rebreathing methods, 3) to assess the effects of cognitive factors on the subjective and physiologic response to CO2, 4) to relate the response to CO2 inhalation and to room air hyperventilation in the laboratory to actual patterns of respiration in patients with panic disorder using ambulatory monitoring, and 5) to study the effects of biological and psychological treatments of panic disorder by repeating CO2 studies following treatment with imipramine, cognitive/behavioral therapy, the combination of the two or placebo. Eighty panic patients, 40 normal controls, and 40 social phobics will undergo CO2 sensitivity assessment with the rebreathing method, a five period canopy study assessing response to 5 and 7% CO2 and to room air hyperventilation, and 24 hour period of ambulatory monitoring. Variation of cognitive set will be employed during the tests in order to assess both psychological as well as biological influences on the generation of panic. Following 12 weeks of treatment testing is repeated.